Trained with 30 images using DPM++_2M for SD3.5L Base model for 6000 steps over 20 epochs//---//Keyword for this LoRa is "Brian Sum". See how the keyword is used in the training prompts here : <a target="_blank" rel="ugc" href="https://imgur.com/a/blPjv6S">https://imgur.com/a/blPjv6S</a>The loss chart + training images can also be found here. Feel free to use this set to train your own LoRa sets.You can download the training data as a zip file here: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/codeShare/lora-training-data">https://huggingface.co/datasets/codeShare/lora-training-data</a>//---//Captions were created using JoyCaption Alpha One : <a target="_blank" rel="ugc" href="https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one">https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one</a>JoyCaption Notebook: <a target="_blank" rel="ugc" href="https://colab.research.google.com/github/camenduru/joy-caption-jupyter/blob/main/joy_caption_jupyter.ipynb">https://colab.research.google.com/github/camenduru/joy-caption-jupyter/blob/main/joy_caption_jupyter.ipynb</a>Keyword originThe keyword "Brian Sum" was selected by running a training image through this notebook: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/codeShare/fusion-t2i-generator-data/blob/main/Google%20Colab%20Jupyter%20Notebooks/fusion_t2i_CLIP_interrogator.ipynb">https://huggingface.co/datasets/codeShare/fusion-t2i-generator-data/blob/main/Google%20Colab%20Jupyter%20Notebooks/fusion_t2i_CLIP_interrogator.ipynb</a>One of the similiar results according to CLIP model in the text_encoding notebook was "art by Brian Sum", so I googled that and behold "Brian Sum" was actually a guy who draws robots! You can find his creations here: <a target="_blank" rel="ugc" href="https://www.artstation.com/sum">https://www.artstation.com/sum .</a>I did add 4 images of Brian Sum's works into the robot LoRa, bringing up the total from 26 images to 30.

Brian Sum Mech LoRa SD3.5 Large

Trained on SD3.5 Large base epoch 17 Loss rate: 0.121 Link to loss chart + training data:<a target="_blank" rel="ugc" href="https://imgur.com/a/blPjv6S">https://imgur.com/a/blPjv6S</a>You can download the training data as a zip file here: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/codeShare/lora-training-data">https://huggingface.co/datasets/codeShare/lora-training-data</a>Captions were created using JoyCaption Alpha One : <a target="_blank" rel="ugc" href="https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one">https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one</a>

mech_art_by_brian_sum_v1.safetensors

v1.0

Trained with 30 images using DPM++_2M for SD3.5L Base model for 6000 steps over 20 epochsRecommended LoRa strength: 0.5CLIP skip: 1Link to loss chart + training data:<a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://imgur.com/a/blPjv6S">https://imgur.com/a/blPjv6S</a>You can download the training data as a zip file here: <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://huggingface.co/datasets/codeShare/lora-training-data">https://huggingface.co/datasets/codeShare/lora-training-data</a>Captions were created using JoyCaption Alpha One : <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one">https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-one</a>JoyCaption Notebook: <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://colab.research.google.com/github/camenduru/joy-caption-jupyter/blob/main/joy_caption_jupyter.ipynb">https://colab.research.google.com/github/camenduru/joy-caption-jupyter/blob/main/joy_caption_jupyter.ipynb</a>Keyword originThe keyword was selected by running a training image through this notebook: <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://huggingface.co/datasets/codeShare/fusion-t2i-generator-data/blob/main/Google%20Colab%20Jupyter%20Notebooks/fusion_t2i_CLIP_interrogator.ipynb">https://huggingface.co/datasets/codeShare/fusion-t2i-generator-data/blob/main/Google%20Colab%20Jupyter%20Notebooks/fusion_t2i_CLIP_interrogator.ipynb</a>One of the similiar results according to CLIP model in the text_encoding notebook was "art by Brian Sum", so I googled that and behold "Brian Sum" was actually a guy who draws robots! You can find his creations here: <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://www.artstation.com/sum">https://www.artstation.com/sum .</a>I did add 4 images of Brian Sum's works into the robot LoRa, bringing up the total from 26 images to 30.

Files with SHA256: 45f833cff3001d61b55209251f5d08ae81e6650d2db9c2a042c47a05f6407fc6